The Put eaux Hoa rd JOHN SILLS n November, 1950, 53 large flan staters of the shows there are two parallel sequences, a ‘normal’ IParisii and a Gallo-Belgic Xb ‘bullet’ stater were one with two or three reverses paired with each brought in to the Bibliothèque Nationale in Paris obverse, and a highly unusual one with at least for identification. Gabrielle Fabre, curator in the eight obverses , three of which also appear in the Cabinet des Médailles, took casts of them and main sequence, paired with a single reverse, my established that they were from a hoard of up to die no. 29 (Sills, 2003, p. 279, fig. 97): the reverse 120 staters found during roadworks in the suburb die usually wore out well before the obverse, and of Puteaux, on the left bank of the Seine near La the pairing of so many obverses with just one Défense. Fabre died at a tragically young age in reverse is wholly exceptional in Gaulish coinage. 1960 leaving only the briefest note describing the Reverse 29, in contrast to all others in the series, circumstances of the find (1951, p. 69), and it fell was deliberately kept in use well beyond the point to Monique Mainjonet to publish the hoard in that it would normally have been discarded, with Revue Numismatique (1962, p. 59-72, pls. 2-5). the result that many coins struck from it are Colbert de Beaulieu later incorporated her work heavily flawed (fig. 2); the coin featured in this list into his general synthesis on the coinage of the (no. 7) shows the die at an early stage, when it had Parisii (1970, p. 20-37); he divided the series as a only just begun to crack (fig. 1). The most likely whole into seven classes, and suggested that the explanation is that a second workshop opened to Puteaux type, class 5, was struck a few years strike a large batch of staters, probably to make an before the start of the Gallic Wars in 58 BC ( ibid , emergency payment, and that quality control was p. 115). In a review article Simone Scheers temporarily sacrificed to speed up production. emphasized the ground-breaking nature of Some idea of the relative size of the batch can be Colbert’s work, but could not resist illustrating got from the number of surviving coins from around fifty coins from major collections and reverse 29, equal to that of the other thirteen class auctions that he had omitted (1972, p. 174-183). 5 reverses combined! The key to understanding this odd phenomenon may lie in two 19 th century hoards discovered while dredging the Seine at Charenton, at the confluence of the Seine and Marne. The Charenton finds, which include class 5 staters, nearly all have test cuts, and this along with their location well to the Fig. 1 Class 5 stater (no. 7 in this list) east of other finds from Paris suggests that they represent payments made to an outside group who saw them essentially as bullion to divide up and melt down, and not as circulating currency. This fits in precisely with what was happening at the mint, where not only had an emergency batch been struck to make a specific payment or payments, but production had also switched between classes 4 and Fig. 2 Class 5 stater, flawed reverse 5 from a standard medium flan stater to a much The Puteaux coins show how important it is to do more impressive broad flan type. Whoever was a full die study of a series and to analyse it camped at Charenton wanted paying in big, brash correctly. Colbert published a reasonably accurate coins: they were suspicious of their paymasters, die chain for class 5 staters, but interpreted it as a hence the test cuts, and didn’t care too much about single sequence (1970, p. 35). Detailed analysis the quality of striking as long as they were good with the benefit of more examples, however, gold, hence the heavily flawed die. The Parisian series can now be tied to the relative either payments made to Balkan and other chronology of Gallo-Belgic Aa and Ca, the broad mercenaries stationed outside the city or money and small flan coinage of the Ambiani or paid directly to German raiders; Puteaux and a Atrebates, and it looks as if class 5 staters are smaller find from boulevard Raspail, closer to the contemporary with Ca class 3 and must therefore city and with no test cuts, may reflect the burial be much earlier than Colbert thought: the best of wealth by the local population. All are precisely guess is that they date from the late 2nd rather contemporary and could have been deposited in than the mid 1 st century. This raises the or around 103, giving us the only fixed point in intriguing possibility that the exceptional features Gaulish numismatics before the Gallic Wars. of the large flan type are related in some way to the invasions of the Cimbri and Teutoni, who plundered Gaul in the last decade of the 2 nd century before being defeated, after several attempts, by Rome. If this was the only evidence we might think that the class 5 emergency batch was struck to pay off Germanic tribes who were threatening the city – and this is still entirely possible – but two extraordinary discoveries from the Balkans suggest a very different explanation. In 1983 a lead weight was found at Prahovo, on the Danube in Eastern Serbia, which reproduces exactly the design of a large flan Parisian stater on one side and has a Latin inscription on the other saying that it had been made by or at the behest of a Roman, Lucius Cestinus, for the king of the Scordisci, the largest Celtic tribe of the region. A Map of Paris showing the places mentioned in the text little later, apparently at nearby Kladovo, a gold Recent archaeological discoveries have thrown stater was discovered with an identical Parisian- more light on the location of the Puteaux hoard type reverse, the only known gold coin of an and a cache of earlier coins found at otherwise silver-using tribe. What on earth was Gennevilliers, in the same loop of the Seine. The going on? One answer is that the Parisii, a wealthy Nanterre district, adjacent to Puteaux, has but relatively small tribe, may have been forced to produced evidence of a 40-acre Iron Age planned employ mercenaries from Celtic tribes in the settlement predating the Gallo-Roman town Balkans and elsewhere to defend themselves from around the Ile de la Cité. Caesar described the Cimbri and Teutoni: the Scordisci had a long Lutetia, the capital of the Parisii, as being sited ‘on tradition of mercenary service. Rome herself may an island in the river Seine’ ( de Bello Gallico 7.57) even have been involved in the transaction, for as and it was long thought that it must have been on well as the legend on the weight we know that in the Ile de la Cité or the Ile St-Louis, despite the 104 BC, after yet another German victory ‘made lack of pre-Roman finds from either. His all Italy tremble with terror’, the Senate broke description by no means contradicts the new with tradition and allowed the recruitment of evidence, however: it is quite possible that the foreign mercenaries; in this febrile atmosphere it main settlement was at Nanterre but that the would have been very much in her interests to population retreated to fortified islets at times of help any Gaulish tribes who were looking to conflict. Single finds of Parisian gold coins are recruit their own mercenaries. In 103 the Cimbri concentrated around the Seine islands, and some and Teutoni are known to have joined forces of these may represent distress hoarding by somewhere north of Parisian territory before their individual households; certainly if there was a last, long march south, and the combined horde mercenary encampment at Charenton it would would have been unstoppable without outside have been ideally placed to guard the islands, but help. So the Charenton hoards seem to represent not Nanterre. Why were none of the four late 2 nd century Bibliography hoards from Paris recovered? We must face the Colbert de Beaulieu, J.-C. 1970: Les monnaies gauloises des possibility that the Parisii and their mercenaries Parisii (Paris). were simply overwhelmed by the massed Fabre, G. 1951: [Communication]. Bulletin de la Société Cimbri and Teutoni. Posidonius, writing around française de Numismatique 6, no. 9, November, p. 69-70. 80 BC, says that the Scordisci ‘have forsworn gold Mainjonet, M. 1962: Le trésor de Puteaux. Revue as an abomination, and will not have it in Numismatique , p. 59-72, pls. 2-5. Scheers, S. 1972: Le monnayage d’or du peuple des Parisii. their country, because of the many terrible things Revue des Archéologues et Historiens d’Art de Louvain 5, p. 174- 183. they suffered because of it’ (Athenaeus, Sills, J.A. 2003: Gaulish and early British gold coinage (London). Deipnosophistae 234a-b). Does this otherwise inexplicable passage refer, in part, to a mercenary Dr John Sills is author of Gaulish and early British gold army that was wiped out on the banks of the coinage (Spink 2003), “the most important work on Celtic Seine? coinage for 30 years.” Chris Rudd List 78, 2004..